- Created a prediction to prredict customer that interested in product ansurance
- Build a model to predict whether policyholders (customers) from last year will also be interested in the Vehicle Insurance provided by the company.
- Find EDA in datasets to get insight valuable data
- Plan a communication strategy to reach those customers and optimize the business model.
- Focused to Purchase Rate
A company that engaged in the insurance sector. Currently, we already have health insurance products and will expand to vehicle insurance.
In this company, we have a Data Science team who asked to build a model to predict whether policyholders (customers) from last year will also be interested in the Vehicle Insurance provided by the company. The number of vehicle grows linearly with its incident numbers
Source : https://www.bps.go.id/indicator/17/513/1/jumlah-kecelakaan-korban-mati-luka-berat-luka-ringan-dan-kerugian-materi.html https://www.bps.go.id/indicator/17/57/1/jumlah-kendaraan-bermotor.html
There are not missing values and duplicate data, and columns for gender, id, region code are drop because to decreased ML Bias
for outlier used IQR method and this is before after outlier
Feature encoding using one hot encoding
Vehicle age that more than 2 years is the rarest fact. and most of people are not previously insured
Random forest outperformed all other ML models by AUC and F-1 score
and for feature importance vehicle damage and previously insured are the top strongest factors
Insight : Highest interest in product among age 41-50 years at 4.17%.
Business Recommendation : Main focused in age range 41-50 for product.
Insight : Highest positive response customers with a Vehicle Age of 1-2 years at 9.10% and Damage Vehicle at 11.92%
Business Recommendation : Gain more customers in Vehicle Age 1-2 years Gain more customers in Vehicle Damage = Yes
Insight : High percentage positive response for Annual Premium in range 20,000-40,000.
Business Recommendation : Focused product rate for Vehicle Insurance in range 20,000 – 40,000.
Insight : Rendahnya ketertarikan pada produk di kalangan usia muda (20-30) Ketertarikan pada kalangan 41-50 pada produk lebih baik
Solution Recommendation : Meningkatkan ketertarikan terhadap asuransi kendaraan bagi kalangan usia muda (20-30 tahun).
Marketing Plan & Strategy : Memanfaatkan social media sebagai sarana dalam mempromosikan produk dan juga mengedukasi orang-orang akan pentingnya mengasuransikan kendaraannya.